Bankruptcy Prediction in Social Enterprises

Traditional bankruptcy literature focuses on commercial enterprises for identifying the strongest variables and models to predict the bankruptcy outcomes. In this study, for the first time, we exploit a large dataset of European bankrupt and healthy social enterprises (SE's) in order to identify the crucial factors that affect the survival of this growing and distinguishable legal form. Combined with the goal of achieving optimal predictive accuracy, we rely on Random Utility Models (RUM) emphasising a new methodology: the Bootstrap Mixed Logit (BMXL). In contrast to what has been found for commercial enterprises, empirical results here indicate that certain organisational features such as the board and workforce size may have a different impact on the probability of SE's bankruptcy.